Photocatalytic electron-transfer oxidation of triphenylphosphine and benzylamine with molecular oxygen via formation of radical cations and superoxide ion

Article abstract of DOI:10.1246/bcsj.79.1489

Photooxygenation of triphenylphosphine (Ph₃P) to triphenylphosphine oxide (Ph₃P=O) with molecular oxygen (O ₂) occurs under photoirradiation of 9-mesityl-10-methylacridinium perchlorate ([Acr⁺-Mes]ClO₄^-) which acts as an efficient electron-transfer photocatalyst. Photooxidation of benzylamine (PhCH₂NH₂) with O₂ also occurs efficiently under photoirradiation of Acr⁺-Mes to yield PhCH₂N=CHPh and hydrogen peroxide (H₂O₂). Each photocatalytic reaction is initiated by intramolecular photoinduced electron transfer from the Mes moiety to the singlet excited state of the Acr⁺ moiety to produce the electron-transfer state (Acr^?-Mes^?+). The Mes^?+ moiety oxidizes Ph₃P and PhCH ₂NH₂ to produce the radical cations (Ph₃P ^?+ and PhCH₂NH₂^?+, respectively), whereas the Acr^? moiety reduces O₂ to O₂^?+. The produced Ph₃P^?+ binds with O₂^?+ as well as O₂, leading to the oxygenated product (Ph₃P=O). On the other hand, proton transfer from PhCH₂NH₂^?+ to O₂ ^?+ occurs, followed by hydrogen transfer, leading to the dehydrogenated dimer product, PhCH₂N=CHPh. In each case, the radical intermediates were detected by laser flash photolysis and ESR measurements to clarify the photocatalytic mechanism.

Full text of DOI:10.1246/bcsj.79.1489

ꢀ
2006 The Chemical Society of Japan
Bull. Chem. Soc. Jpn. Vol. 79, No. 10, 1489–1500 (2006) 1489
BCSJ Award Article
Photocatalytic Electron-Transfer Oxidation of Triphenylphosphine
and Benzylamine with Molecular Oxygen via Formation
of Radical Cations and Superoxide Ion
ꢀ
Kei Ohkubo, Takashi Nanjo, and Shunichi Fukuzumi
Department of Material and Life Science, Graduate School of Engineering, Osaka University, Suita, Osaka 565-0871
SORST, Japan Science and Technology Agency (JST), Suita, Osaka 565-0871
Received April 20, 2006; E-mail: fukuzumi@chem.eng.osaka-u.ac.jp
Photooxygenation of triphenylphosphine (Ph3P) to triphenylphosphine oxide (Ph3P=O) with molecular oxygen
þ ꢁ
(O2) occurs under photoirradiation of 9-mesityl-10-methylacridinium perchlorate ([Acr –Mes]ClO4 ) which acts as
an efficient electron-transfer photocatalyst. Photooxidation of benzylamine (PhCH2NH2) with O2 also occurs efficiently
þ
under photoirradiation of Acr –Mes to yield PhCH2N=CHPh and hydrogen peroxide (H2O2). Each photocatalytic re-
action is initiated by intramolecular photoinduced electron transfer from the Mes moiety to the singlet excited state of the
þ
ꢂ
ꢂþ
ꢂþ
Acr moiety to produce the electron-transfer state (Acr –Mes ). The Mes moiety oxidizes Ph3P and PhCH2NH2 to
ꢂþ
ꢂþ
ꢂ
ꢂꢁ
produce the radical cations (Ph3P and PhCH2NH2 , respectively), whereas the Acr moiety reduces O2 to O2 . The
produced Ph3P binds with O2 as well as O2, leading to the oxygenated product (Ph3P=O). On the other hand, proton
transfer from PhCH2NH2 to O2 occurs, followed by hydrogen transfer, leading to the dehydrogenated dimer prod-
uct, PhCH2N=CHPh. In each case, the radical intermediates were detected by laser flash photolysis and ESR measure-
ments to clarify the photocatalytic mechanism.
ꢂþ
ꢂꢁ
ꢂþ
ꢂꢁ
Free radicals are highly versatile and reactive intermediates,
allowing many syntheses to be carried out under relatively
reactions of O2 are known to be rather slow because of the
7–31
large reorganization energy of electron transfer.2
case of O2, however, the excited state (singlet oxygen) is fre-
In the
1–4
mild conditions in chemical and biological systems. How-
ever, the high reactivity of free radicals usually results in
low selectivity and thus, it has been difficult to utilize reactions
between free radicals for selective bond formation.1 In con-
trast to neutral free radicals, reactions of radical ions are well
controlled by charges because radical cations are expected to
react selectively with radical anions as compared with reac-
tions of radical ions with the same charge. Thus, reactions be-
tween radical cations and radical anions have profound funda-
quently formed in reactions involving photoexcited states via
Thus, it
energy transfer, reacting directly with substrates.3
2–35
is highly desired to avoid such an energy-transfer process for
–4
ꢂꢁ
the selective reaction between radical cations and O2
We have recently reported that the photoexcitation of 9-
.
þ
mesityl-10-methylacridinium ion (Acr –Mes) results in ultra-
fast electron transfer from the Mes moiety to the singlet excit-
þ
ed state of the Acr moiety to form the electron-transfer state
ꢂ ꢂþ
mental and synthetic interest for selective bond formation.5
–11
(Acr –Mes ), which has an extremely long lifetime (e.g., 2 h
at 203 K) and a high energy (2.37 eV).
36,37
Radical cations and radical anions can be produced photo-
chemically by electron transfer from photoexcited electron do-
The electron-trans-
fer state (Acr –Mes ) is both a strong oxidizing agent with
ꢂ
ꢂþ
ꢀ
ꢀ
ꢂþ
nors (D : denotes the photoexcited state) to electron accep-
tors (A) or from electron donors (D) to photoexcited electron
the Mes moiety and the a strong reducing agent with the
ꢂ
36,37
þ
Acr moiety. In such a case, Acr –Mes acts as an efficient
electron-transfer photocatalyst for highly selective oxygena-
ꢀ
by photoexcitation of the charge-transfer complex between D
12–17
acceptors (A ).
Radical ion pairs can also be produced
tion of anthracene and tetraphenylethylene with O2 via selec-
ꢂꢁ
1
8–26
ꢂþ
ꢂþ
ꢂꢁ
and A.
sult in back electron transfer from A to D to reproduce the
reactant pair (D and A) without the formation of a chemical
However, reactions between D and A often re-
ꢂꢁ
tive radical coupling of the donor radical cations and O2 un-
der visible light irradiation to produce epidioxyanthracene and
1,2-dioxetane, respectively.3
8,39
However, the use of Acr –
þ
ꢂꢁ
ꢂþ
bond. This back electron transfer from A to D can be re-
tarded by choosing an appropriate substance, which has large
reorganization energy for electron transfer. Molecular oxygen
Mes as an efficient electron-transfer photocatalyst has yet to
be expanded to photooxidation of substrates with O2 other than
cycloadditions.
þ
(O2) is suitable for such a purpose, because electron-transfer
Herein, we expand the scope of the use of Acr –Mes as an
Published on the web October 10, 2006; doi:10.1246/bcsj.79.1489

1
490 Bull. Chem. Soc. Jpn. Vol. 79, No. 10 (2006) BCSJ AWARD ARTICLE
efficient electron-transfer photocatalyst for oxidation of organ-
ic compounds with O2 other than oxygenation reactions by ex-
amining photocatalytic oxidation of triphenylphosphine and
(s, 2H), 7.22–7.32 (m, 1H), 7.33–7.38 (d, J ¼ 4:5 Hz, 4H), 7.41–
7.48 (m, 3H), 7.74–7.81 (m, 2H).
The amount of H2O2 was determined by titration by iodide ion
as follows.4
3–45
þ
2
An O -saturated MeCN solution (3.0 mL) of
benzylamine with O2 in the presence of Acr –Mes. The pho-
þ
ꢁ4
ꢁ3
ꢁ4
Acr –Mes (1:0 ꢃ 10 mol dm ) and benzylamine (1:0 ꢃ 10
tocatalytic mechanisms are clarified by detecting radical inter-
mediates involved in the photocatalytic reactions with use of
laser flash photolysis and ESR measurements.
ꢁ3
mol dm ) was photoirradiated for 2 h and the photoirradiated
product mixture was treated with excess amount of NaI. The
ꢁ
amount of I3 formed was then determined from the UV–vis spec-
trum (ꢁmax ¼ 361 nm, " ¼ 2:50 ꢃ 10 mol dm cm ).44,45
Quantum Yield Determinations. A standard actinometer
[potassium trisoxalatoferrate(III)]46 was used to determine the
quantum yield of the photocatalytic oxygenation of Ph3P with
4
ꢁ1
3
ꢁ1
Experimental
Materials. Benzylamine, N,N-dimethylbenzylamine, N-benzyl-
idenebenzylamine, and triphenylphosphine oxide (Ph3P=O) were
obtained from Tokyo Kasei Kogyo Co., Ltd. Triphenylphosphine
þ
O2 in the presence Acr –Mes. Typically, a square quartz cuvette
(10 mm i.d.), which contained a degassed MeCN or CHCl3 solution
(
Ph3P) was obtained from Sigma-Aldrich Co., Ltd. Di-t-butylper-
t
t
3
þ
ꢁ4
ꢁ3
oxide (Bu OOBu ) was obtained from Nakalai Tesque Co., Ltd.
Acetonitrile (MeCN) and chloroform (CHCl3) were of spectral
grade, obtained commercially and used without further purification.
(3.0 cm ) of Acr –Mes (2:0 ꢃ 10 mol dm ) and Ph3P (2:0 ꢃ
ꢁ4
ꢁ3
10 mol dm ), was irradiated with steady-state monochromat-
ized light (ꢁ ¼ 430 nm) from a Shimadzu RF-5300PC fluo-
rescence spectrophotometer. Under the conditions of actinometry
þ ꢁ
9
-Mesityl-10-methylacridinium perchlorate ([Acr –Mes]ClO4 )
þ
was prepared by the reaction of 10-methylacridone with mesityl
magnesium bromide (MesMgBr) in tetrahydrofuran, followed by
addition of sodium hydroxide (water) for hydrolysis and perchlo-
ric acid for neutralization.38 The compound was purified by re-
crystallization from methanol–diethyl ether. The yield was 87%
based on acridone.40 H NMR (CD3CN, 300 MHz) ꢀ (ppm) 8.16
experiments, the actinometer and Acr –Mes absorbed essentially
all of incident light of ꢁ ¼ 430 nm. The light intensity of mono-
ꢁ
7
chromatized light of ꢁ ¼ 430 nm was determined as 9:9 ꢃ 10
ꢁ3
ꢁ1
mol dm
which contained a degassed CHCl3 solution (3.0 cm ) of Acr –
s
. In addition, a square quartz cuvette (10 mm i.d.),
3
þ
1
ꢁ4
ꢁ3
ꢁ4
ꢁ3
Mes (2:0 ꢃ 10 mol dm ) and Ph3P (2:0 ꢃ 10 mol dm ),
was irradiated with monochromatized laser flash (ꢁ ¼ 430 nm)
from Nd:YAG laser (Continuum, SLII-10, 4–6 ns fwhm) at 430
nm. Under the conditions of actinometry experiments, the actino-
(
(
d, J ¼ 9:0 Hz, 2H), 7.93 (t, J ¼ 9:0 Hz, 2H), 7.40 (s, 4H), 6.79
s, 2H), 4.37 (s, 3H), 2.02 (s, 3H), 1.25 (s, 6H). Anal. Calcd for
C23H22ClNO4: C, 67.07; H, 5.38; N, 3.40%. Found: C, 66.78;
þ ꢁ
þ
H, 5.33; N, 3.35%. 10-Methylacridinium iodide (AcrH I ) was
prepared by the reaction of acridine with methyl iodide in acetone
meter and Acr –Mes absorbed essentially all of incident light of
ꢁ ¼ 430 nm. The light intensity of monochromatized light of ꢁ ¼
þ
ꢁ
ꢁ8
and was converted to the perchlorate salt (AcrH ClO4 ) by the
þ ꢁ
430 nm was determined as 8:6 ꢃ 10 einstein/pulse. The photo-
addition of magnesium perchlorate to the iodide salt (AcrH I ),
4
chemical reaction was monitored using a Hewlett Packard 8453
diode-array spectrophotometer. The quantum yields were deter-
mined from a decrease in absorbance due to Ph3P (ꢁ ¼ 302 nm,
1
and purified by recrystallization from methanol. Synthesis of
2
[
N,N- H2]benzylamine was carried out according to the litera-
4
2
3
ꢁ1
3
ꢁ1
ture. A solution containing benzylamine (10 g) in CH2Cl2 was
washed thrice with 10 mL portions of D2O. The solution contain-
ing the deuterated product was dried over K2CO3, and the solvent
" ¼ 1:69 ꢃ 10 mol dm cm ) and benzylamine (ꢁ ¼ 309 nm,
2 3
ꢁ1
ꢁ1
" ¼ 2:7 ꢃ 10 mol dm cm ). In order to avoid the contribution
by light absorption of the products, only the initial rates were used
for determination of the quantum yields.
1
was removed in vacuo. From the H NMR spectrum, there was
>
90% deuterium incorporation.
Reaction Procedures. The photocatalytic oxygenation of
Ph3P with O2 was carried out by the following procedure. Typical-
Fluorescence Quenching. Quenching experiments of the fluo-
rescence of photocatalysts by Ph3P or benzylamine were per-
formed using a Shimadzu RF-5300PC fluorescence spectropho-
tometer. The monitoring wavelength was the maximum of the
þ
ly, a chloroform solution (2 mL) containing Acr –Mes (4.12 mg,
1
:0 ꢃ 10ꢁ mol) and Ph3P (7.87 mg, 3:0 ꢃ 10 mol) in a Schlenk
5
ꢁ5
emission band for AcrH . The solutions were deoxygenated by
argon purging for 10 min prior to the measurements. Relative
þ
flask with a rubber septum was saturated with oxygen by bubbling
oxygen through a stainless steel needle for 20 min. The solution
was then irradiated with a 500 W xenon lamp (Ushio Optical
ModelX SX-UID 500XAMQ) through a color filter glass (Asahi
Techno Glass Y43) transmitting ꢁ > 430 nm at 278 K. After
emission intensities were measured for MeCN or CHCl3 solutions
þ
ꢁ6
ꢁ3
containing AcrH (5:0 ꢃ 10 mol dm ) with electron donors at
ꢁ2
ꢁ3
various concentrations (0{8:0 ꢃ 10 mol dm ). There was no
change in the shape, but there was a change in the intensity of
the fluorescence spectrum by the addition of an electron donor.
The Stern–Volmer relationship (Eq. 1) was obtained for
30 min photoirradiation, the corresponding Ph3P=O was identified
by comparison of the GC retention time and the GC/MS spectrum
(
Shimadzu QP-5000) with those of an authentic sample.
The photocatalytic oxidation of benzylamine with O2 was car-
I0=I ¼ 1 þ KSV½Dꢄ;
the ratio of the emission intensities in the absence and presence of
ð1Þ
ried out as follows. Typically, an acetonitrile-d3 solution (0.7 mL)
þ
ꢁ6
containing Acr –Mes (1.44 mg, 3:5 ꢃ 10 mol) and benzylamine
electron donor (I0=I) and concentrations of quenchers [D]. The
ꢁ5
þ
(
4.87 mg, 3:8 ꢃ 10 mol) in a NMR tube with a rubber septum
fluorescence lifetimes (ꢂ) are 37 ns for AcrH and 31 ns in MeCN
was saturated with oxygen by bubbling oxygen gas through a
stainless steel needle for 20 min. The solution was then irradiated
with a 500 W xenon lamp (Ushio Optical ModelX SX-UID
and CHCl3, respectively.47 The fluorescence lifetimes of AcrH
þ
were measured by a Photon Technology International GL-3300
with a Photon Technology International GL-302, nitrogen laser/
pumped dye laser system, equipped with a four channel digital
delay/pulse generator (Stanford Research System Inc. DG535)
and a motor driver (Photon Technology International MD-5020).
Excitation wavelength was 430 nm using a toluene solution con-
taining dimethyl-POPOP [1,4-bis(4-methyl-5-phenyl-2-oxazolyl)-
5
00XAMQ) through a color filter glass (Asahi Techno Glass
Y43) transmitting ꢁ > 430 nm at 278 K. After photoirradiation for
h, N-benzylidenebenzylamine was identified by comparing the
6
1
H NMR spectrum in comparison with that of an authentic sample.
1
N-Benzylidenebenzylamine: H NMR (300 MHz, CD3CN) ꢀ 4.77

K. Ohkubo et al.
Bull. Chem. Soc. Jpn. Vol. 79, No. 10 (2006) 1491
benzene] (Dojindo, Japan) as a laser dye. The observed quench-
ꢁ
formation of an oxygenated product, i.e., triphenylphosphine
52,53
oxide (Ph P=O) in a 99% yield (Eq. 2).
ing rate constants kobs (¼KSV
ꢂ
1) were obtained from the Stern–
3
Volmer constants KSV and the emission lifetimes ꢂ.
Electrochemical Measurements. Second harmonic ac vol-
tammetry (SHACV)48,49 measurements of Ph3P, benzylamine,
and N-benzylidenebenzylamine were performed with an ALS630B
hν
2 Ph3P
+
O2
2 Ph P=O
3
ð2Þ
Me
electrochemical analyzer in degassed acetonitrile containing 0.1
ꢁ
mol dm 3 Bu4N ClO4 (TBAP) as a supporting electrolyte or in
ꢁ
þ
Me
Me
ꢁ
3
chloroform containing 0.2 mol dm TBAP at 298 K. The platinum
working electrode (BAS, surface i.d. 1.6 mm) was polished with
BAS polishing alumina suspension and rinsed with acetone before
use. The counter electrode was a platinum wire (0.5 mm dia.). The
measured potentials were recorded with respect to an Ag/AgNO3
N
Me
+
Acr −Mes
(
0.01 mol dm 3) reference electrode. The values of redox poten-
ꢁ
þ
tials (vs Ag/AgNO3) were converted into those vs SCE by addition
of 0.29 V.50
ESR Measurements. ESR measurements were performed
on a JEOL X-band ESR spectrometer (JES-ME-LX) at 233 K. A
quartz ESR tube (internal diameter: 1.5 mm) containing an argon-
When Acr –Mes was replaced by 10-methylacridinium ion
þ
(AcrH ), Ph3P=O was also obtained quantitatively after 60
min photoirradiation. It was confirmed that no photooxygena-
þ
þ
tion of Ph P occurred without Acr –Mes or AcrH under
3
otherwise the same experimental conditions.
The quantum yields (ꢀP) of formation of Ph3P=O under
photoirradiation with monochromatized light (ꢁ ¼ 430 nm)
were determined from a decrease in absorbance due to Ph3P
ꢁ
1
saturated di-t-butylperoxide solution of benzylamine (3:0 ꢃ 10
ꢁ
3
mol dm ) at 243 K was irradiated in the cavity of the ESR spec-
trometer with the focused light of a 1000-W high-pressure Hg lamp
(
Ushio-USH1005D) through an aqueous filter. The ESR spectra
using an actinometer (see Experimental). The dependence of
þ
were measured under nonsaturating microwave power conditions.
The amplitude of modulation was chosen to optimize the resolution
and the signal-to-noise (S=N) ratio of the observed spectra.
Laser Flash Photolysis Measurements. Measurements of
transient absorption spectra in the photochemical reactions of
ꢀ
P value on concentration of Ph3P in the AcrH -catalyzed
photooxygenation of Ph3P with O2 in air-saturated CHCl3 is
shown in Fig. 1a (closed square). The ꢀP value increases with
increasing concentration of Ph3P [Ph3P] to reach a constant
value. The saturated dependence of ꢀP on [Ph3P] was convert-
þ
Acr –Mes with Ph3P and benzylamine were performed according
to the following procedure. A degassed CH3CN3 solution contain-
ꢁ1
ꢁ1
ed to a linear plot of ꢀP vs [Ph3P] as shown in Fig. 1b.
þ
ꢁ5
ꢁ3
From the slope and intercept, the ꢀ1 and Kobs values were
ing Acr –Mes (8:0 ꢃ 10 mol dm ) and benzylamine (1:3 ꢃ
2
ꢁ1
3
ꢁ ꢁ3
2
determined to be 0.57 and 2:1 ꢃ 10 mol dm , respectively.
1
0
mol dm ) was excited by Nd:YAG laser (Continuum,
The Kobs value is converted to rate constant (kobs) using the re-
SLII-10, 4–6 ns fwhm) at 430 nm. Time courses of the transient
absorption spectra were measured by using a continuous Xe-lamp
ꢁ
þ
1
lation: kobs ¼ Kobs
ꢂ
cited state of AcrH ( AcrH : 31 ns) provided that AcrH
where ꢂ is the lifetime of the singlet ex-
1
þꢀ
1
þꢀ
(
150 W) and an In GaAs-PIN photodiode (Hamamatsu 2949) as a
47
reacts with Ph3P in the photocatalytic reaction. The kobs val-
probe light and a detector, respectively. The output from the pho-
todiodes and a photomultiplier tube was recorded with a digitizing
oscilloscope (Tektronix, TDS3032, 300 MHz). The transient spec-
tra were recorded using fresh solutions in each laser excitation.
The measurements of transient absorption spectra in the photo-
ue was determined to be 6:8 ꢃ 10 mol dm3 s . The kobs
9
ꢁ1
ꢁ1
value was also obtained from the fluorescence quenching of
þꢀ
1
AcrH
1
by electron transfer from an electron donor (Ph3P)
9
þꢀ
ꢁ1
3
ꢁ1
to AcrH as 9:0 ꢃ 10 mol dm s (See Supporting Infor-
t
t
9
chemical reaction of Bu OOBu with benzylamine were per-
t
mation S1), which is comparable with the value (9:0 ꢃ 10
t
ꢁ1 ꢁ1
3
formed as follows. Typically, a degassed Bu OOBu solution
mol dm s ) obtained from the dependence of ꢀ on [Ph P]
P 3
ꢁ
1
ꢁ3
þ
containing benzylamine (3:0 ꢃ 10 mol dm ) was excited by
Nd:YAG laser (Continuum, SLII-10, 4–6 ns fwhm) at 355 nm.
All experiments were performed at 298 K.
in Fig. 1a (closed square). This indicates that the AcrH -cata-
lyzed photooxygenation of Ph3P with O2 proceeds via electron
transfer from Ph3P to AcrH
When AcrH was replaced by Acr –Mes, the ꢀP values
remained constant (0.58) irrespective of the concentration of
Ph3P as shown in Fig. 2 (closed circle), in contrast with the
1
þꢀ
.
Theoretical Calculations. Density-functional theory (DFT)
calculations were performed on a COMPAQ DS20E computer.
Geometry optimizations were carried out using the Becke3LYP
þ
þ
functional and 6-31G basis set51 with unrestricted Hartree–Fock
ꢀ
þ
case of the AcrH -catalyzed photooxygenation of Ph3P with
þ
ꢂþ
(UHF) formalism for benzylamine radical cation and Ph3P as
implemented in the Gaussian 03 program. Graphical output of
the computational results were generated with the Cerius soft-
ware program developed by Molecular Simulations Inc.
O2 (Fig. 1a). In the case of Acr –Mes, photoirradition of
þ
Acr –Mes results in formation of the electron-transfer state
þ
2
ꢂ
ꢂþ 36
of Acr –Mes (Acr –Mes ). Even small concentrations of
ꢂ
ꢂþ
ꢂþ
Ph3P can react with the Mes moiety of Acr –Mes by elec-
ꢂ
ꢂþ
þꢀ
tron transfer because the lifetime of Acr –Mes (ca. ms) is
1
Results and Discussion
36,47
extremely long as compared to that of AcrH
(31 ns).
Photocatalytic Oxidation of Triphenylphosphine with
O2. Visible light irradiation (ꢁ > 430 nm) of the absorp-
tion band of 9-mesityl-10-methylacridinium ion (Acr –
This may be the reason why the ꢀP values are constant (0.58)
in the concentration range of Ph3P studied, Fig. 1a.
þ
The dependence of ꢀP on the concentration of O2 at a fixed
ꢁ3
ꢁ3
ꢁ4
ꢁ3
Mes; 5:0 ꢃ 10 mol dm ) in an O2-saturated chloroform
concentration of Ph3P (5:0 ꢃ 10 mol dm ) under photoirra-
diation with monochromatized light (ꢁ ¼ 430 nm) is shown in
Fig. 2 (closed circle). The ꢀP value increases with increasing
(
1
CHCl3) solution containing triphenylphosphine (Ph3P; 1:5 ꢃ
ꢁ ꢁ3
2
0
mol dm ) for 30 min by xenon lamp resulted in the

1
492 Bull. Chem. Soc. Jpn. Vol. 79, No. 10 (2006) BCSJ AWARD ARTICLE
(a)
(b)
0
.8
4.5
4
3
.0
.5
0.6
0.4
0.2
0
3.0
2
.5
.0
2
1.5
0
1
2
3
4
0
0.5
1.0
1.5
2.0
2.5
2
–3
10–2 [Ph P] / mol–1dm3
–1
1
0 [Ph P] / mol dm
3
3
Fig. 1. (a) Dependence of quantum yield (ꢀP) of Ph3P=O on concentrations of Ph3P for the photocatalytic oxygenation of Ph3P
þ
ꢁ5
ꢁ3
þ
ꢁ5
ꢁ3
ꢁ1
with Acr –Mes (8:0 ꢃ 10 mol dm
,
) and AcrH (8:0 ꢃ 10 mol dm
,
) in air-saturated CHCl3. (b) Plot of ꢀP on
ꢁ1
þ
[
Ph3P] for the photocatalytic oxygenation of Ph3P with AcrH in air-saturated CHCl3.
0
0
0
.9
.6
.3
0
electron transfer from the Mes moiety to the singlet excited
36
þ
state of the Acr moiety. The quantum yield for the forma-
ꢂ
ꢂþ
tion of Acr –Mes has been previously determined to be
nearly quantitative (98%). Since the one-electron reduction
36
ꢂ
ꢂþ
0
potential of Acr –Mes (E red ¼ 1:88 V vs SCE) is more pos-
0
itive than the one-electron oxidation potential of Ph3P (E ox ¼
1:20 V vs SCE), determined by second harmonic AC voltam-
metry (SHACV, See Experimental and Supporting Information
ꢂþ
Fig. S2), electron transfer from Ph3P to the Mes moiety in
ꢂþ
ꢂ
Acr –Mes is energetically feasible. Thus, the addition of
þ
Ph3P to a CHCl3 solution of Acr –Mes and laser photoexcita-
ꢂþ
tion afforded the Ph3P radical cation (Ph3P : ꢁmax ¼ 380
nm)5
4,55
as shown in Fig. 3a (open circles). The Acr moiety
ꢂ
36
ꢂ
ꢂþ
of Acr –Mes has an absorption band at 520 nm. In the
presence of Ph3P, electron transfer from Ph3P to the Mes
moiety occurs to produce Ph3P
0
2
4
6
8
10
ꢂþ
3
–3
1
0 [O ] / mol dm
ꢂþ ꢂ
2
and Acr –Mes, both of
which have an absorption band at 520 nm. This is the rea-
son why the absorption at 520 nm becomes larger in the
presence of Ph3P (open circles in Fig. 3a) as compared with
that in the absence of Ph3P (closed circles in Fig. 3b). The
formation and decay of the absorption at 380 nm are the
same as those at 520 nm. It is important to note that the ab-
Fig. 2. Dependences of quantum yield (ꢀP) of Ph3P=O on
concentration of O2 for the photocatalytic oxygenation of
ꢁ4
ꢁ3
þ
ꢁ5
Ph3P (5:0 ꢃ 10 mol dm ) with Acr –Mes (8:0 ꢃ 10
ꢁ
3
mol dm ) photoirradiated by laser flash photolysis (ꢁ ¼
30 nm, 30 mJ/pulse) ( ) and steady-state photolysis with
monochromatized light (ꢁ ¼ 430 nm) ( ).
4
ꢂ
sorption band due to the Acr moiety remains virtually the
same in the absence of O2 (Fig. 3b). This indicates clearly that
concentration of O2 to reaching a maximum at ꢀ1 ¼ 0:78.
When the photoirradiation was carried out by using pulsed
laser (ꢁ ¼ 355 nm, Nd-YAG laser, 10 Hz, 30 mJ/pulse) in-
stead of steady-state photolysis with monochromatized light,
however, the ꢀP values remained constant (0.21) irrespective
of different the O2 concentration. The origin of the different
dependencies of ꢀP on concentration of O2 between the laser
excitation and steady-state photolysis with monochromatized
light is discussed in the following section.
ꢂþ
ꢂþ
ꢂ
Ph3P is formed by electron transfer from Ph3P to Acr –
Mes rather than by direct photoinduced electron transfer
þ
from Ph3P to the singlet excited state of the Acr moiety in
þ
Acr –Mes.
The rate of formation of Ph3P increases with increasing
ꢂþ
concentration of Ph3P as shown in Fig. 4a. The formation rate
obeys pseudo-first-order kinetics, and the pseudo-first-order
rate constant increases linearly with increasing concentration
of Ph3P (Fig. 4b). The second-order rate constant of electron
Photocatalytic Mechanisms of Oxygenation of Ph3P. In
order to elucidate the photocatalytic oxygenation mechanism,
laser flash photolysis was used to detect the intermediates.
Nanosecond laser excitation at 430 nm of a degassed anhy-
ꢂ
ꢂþ
transfer from Ph3P to Acr –Mes was determined from the
9
ꢁ1
3
slope of the linear plot in Fig. 4b to be 9:2 ꢃ 10 mol dm
ꢁ1
s
ed for the exergonic electron transfer.
, which is close to be the diffusion-limited value as expect-
56
þ
drous CHCl3 solution of Acr –Mes resulted in the formation
ꢂ
ꢂþ
of the electron-transfer state (Acr –Mes ) via photoinduced
The observed residual absorption at 380 nm is attributed

K. Ohkubo et al.
Bull. Chem. Soc. Jpn. Vol. 79, No. 10 (2006) 1493
(
a)
(b)
.03
0
0
0
[
Ph P] = 5 x 10–2 mol dm–3 (10
µs)
3
[Ph3P] = 1.5 x 10–2 mol dm–3 (10
µs)
0
.04
.02
0
.02
.01
0
0
[Ph3P] = 0 (10
µs)
–
0.02
–0.01
400
500
600
700
800
400
500
600
700
800
Wavelength / nm
Wavelength / nm
þ
ꢁ5
Fig. 3. Transient absorption spectra observed in photoinduced electron-transfer oxidation of Ph3P with Acr –Mes (8:0 ꢃ 10
ꢁ3
mol dm ) taken 10 ms after laser excitation at 430 nm (a) in deaerated CHCl3 at 298 K. (b) in O2-saturated CHCl3 at 298 K.
(
a)
(b)
0
0
0
0
.04
.03
.02
10
8
1
05 [Ph3P] / M
8
6
4
2
.0
.0
.0
.0
6
4
.01
0
2
0
0
1
2
3
4
5
0
2
4
6
8
10
5
Time / µs
10 [Ph P] / M
3
ꢂþ
Fig. 4. (a) Time profiles at 380 nm of transient absorption due to the formation of Ph3P observed by photoexcitation of a dea-
erated MeCN solution containing Acr –Mes (8:0 ꢃ 10 mol dm ) and various concentrations of Ph3P (2:0 ꢃ 10ꢁ5–8:0 ꢃ 10ꢁ5
þ
ꢁ5
ꢁ3
ꢁ3
mol dm ) with single-exponential curve. (b) Plot of pseudo-first-order rate constant (k1) vs concentration of Ph3P.
to Ph3P (" ¼ 3600 mol dm cm )54 which slowly decays,
Fig. 5. The decay obeys the second-order kinetics as shown in
the inset of Fig. 5. From the slope of the second-order plot, the
ꢂþ
ꢁ1
3
ꢁ1
Ph3PO2 with Ph3P in competition with the dissociation to Ph3P
and O2 (Scheme 1).5
4,57–59
Ph3PO2 may also be formed via
ꢂþ
ꢂþ
addition of O2 to Ph3P to give Ph3PO2 and back electron
1
0
ꢁ1
ꢂ
ꢂþ
decay rate constant was determined to be 1:1 ꢃ 10 mol
transfer from Acr –Mes to Ph3PO2 (reaction pathway B in
58,59
Scheme 1).
3
ꢁ1
dm s , which is close to diffusion rate in CHCl3. This proc-
ꢂꢁ
ess corresponds to the back electron transfer from O2 to
ꢂþ
Based on Scheme 1, the competition between the pathways
ꢂþ
Ph3P and radical coupling to give the oxygenated adduct.
The photocatalytic mechanism involving the radical inter-
mediates detected in Fig. 4 is shown in Scheme 1. Photoexci-
A and B is determined by the rate of the reaction of Ph3P
ꢂꢁ ꢂþ
with O2 and O2 . The rate constant of the reaction of Ph3P
6
ꢁ1
3
ꢁ1 58
ꢂþ
ꢁ1
with O2 (kB) was reported to be 7:3 ꢃ 10 mol dm s
,
þ
tation of Acr –Mes gives the long-lived electron-transfer state
ꢂþ
and the first-order rate constant (k1) of the reaction of Ph3P
ꢂ
ꢂþ
ꢂ
ꢁ3
(
Acr –Mes ). The Acr and Mes moieties can reduce and
ꢂþ
with O2 in O2-saturated CHCl3 (½O2ꢄ ¼ 7:8 ꢃ 10 mol
3
60
4
ꢁ1
oxidize O2 and Ph3P , respectively, to give superoxide anion
dm ) was estimated to be 5:7 ꢃ 10 s . On the other hand,
ꢂꢁ
ꢂþ
ꢂþ
ꢂꢁ
ꢂþ
ꢂꢁ
(
O2 ) and Ph3P , respectively. Radical coupling between
the rate constant of the reaction of Ph3P with O2 (kA)
was determined to be 1:1 ꢃ 10 mol dm3 s (Fig. 5). The
10
ꢁ1
ꢁ1
Ph3P and O2 may afford the dioxygen adduct (Ph3PO2)
reaction pathway A in Scheme 1), which may be a cyclic per-
ꢂþ
(
oxide as reported in the literature. The final oxygenated prod-
uct (Ph3P=O) is known to be formed by the reaction of
concentration of Ph3P formed in electron transfer from Ph3P
57
ꢂ
ꢂþ
ꢂꢁ
to Acr –Mes should be the same as concentration of O2
ꢂ
formed in electron transfer from Acr –Mes to O2 (Scheme 1).

1
494 Bull. Chem. Soc. Jpn. Vol. 79, No. 10 (2006) BCSJ AWARD ARTICLE
·
·+
Ph P
3
Acr −Mes
h
ν
+
O2
kB
·+
·+
Ph3P
Ph PO
3 2
Me
N
Me
Me
(B)
·
Acr −Mes
Ph P + O
3
2
+
Me (Acr −Mes)
(A)
Ph PO
3 2
kA
+
Ph P
3
2 Ph P=O
3
·−
O2
O2
Scheme 1.
1AcrH*
0
0
0
0
.08
.06
.04
.02
0
1
000
+
Ph P
3
h
ν
·
·+
3
AcrH + Ph P
5
00
H
N
Me
+
O2
(AcrH+)
0
0
50
100
Time / µs
150
200
·
·+
2
AcrH + Ph PO
3
Ph PO
3
2
+
Ph P
3
0
100
200
300
Ph P + O
2 Ph P=O
3
3
2
Time / µs
Scheme 2.
Fig. 5. Decay time profile at 380 nm of transient absorption
ꢂþ
due to Ph3P observed by photoexcitation of an oxygen-
þ
ꢂþ
ꢂꢁ
Ph3P and O2 is almost unity. The first-order rate constant
saturated MeCN solution containing Acr –Mes (8:0 ꢃ
ꢂþ
ꢂꢁ
ꢁ9
ꢁ5
ꢁ3
(k1) of the reaction of Ph3P
with O2
(9:1 ꢃ 10 mol
10
mol dm ). Inset: second-order plot.
ꢁ
3
2
ꢁ1
dm ) was to be 1:0 ꢃ 10 s . Thus, the ratio of pathway A
to B under the steady-state photoirradiation with monochrom-
atized light (ꢁ ¼ 430 nm) is approximately 1:570 from the
ꢂþ
The concentration of Ph3P in the laser flash photolysis was
ꢁ3
ꢁ6
determined to be 5:1 ꢃ 10 mol dm
from the transient
absorption of Acr –Mes (ꢁmax ¼ 500 nm and "500 ¼ 3:5 ꢃ
ꢂ
ꢂþ
3
2
ratio of the k1 values (6:0 ꢃ 10 to 1:0 ꢃ 10 ). Ph3PO2 reacts
3
ꢁ1
3
ꢁ1 36
ꢂþ
1
0 mol dm cm
)
in Fig. 4. In this case, the k1 value of
ꢁ3
with the large excess of Ph3P to give Ph3P=O via O–O bond
57
cleavage.
ꢂꢁ
ꢁ6
the reaction of Ph3P with O2 (5:1 ꢃ 10 mol dm ) was
4
ꢁ1
þ
determined to be 5:6 ꢃ 10 s . Thus, the ratio of pathway A
In contrast to the case of Acr –Mes in Scheme 1, the
þ
to B was estimated to be 1:1 from the ratio of the k1 values
4
AcrH -catalyzed photooxygenation of Ph3P with O2 proceeds
4
1
þꢀ
shown in Scheme 2. Photoinduced electron transfer from
(
5:6 ꢃ 10 to 5:7 ꢃ 10 ). This indicates that the photooxygena-
via electron transfer from Ph3P to AcrH (vide supra) as
61
tion proceeds via pathway A as well as pathway B in the case
of laser pulse excitation.
In the case of the steady-state photoirradiation with mono-
1
þꢀ
0
Ph3P to AcrH
is energetically feasible, because the E ox
0
value of Ph3P (1.20 V vs SCE) is lower than the E red value
of AcrH
transfer from Ph3P to AcrH , reacts with O2 to yield tri-
phenylphosphine peroxide radical cation (Ph3PO2 ). Then,
ꢂþ
1
þꢀ ꢂþ
47
chromatized light (ꢁ ¼ 430 nm), the concentration of Ph3P
(2.32 V). Ph3P , produced by the electron
1
ꢂꢁ
þꢀ
and O2 is given by Eq. 4, where In is the light intensity
ꢁ
7
ꢁ3 ꢁ1
10
ꢁ1
ꢂþ
ꢂþ
(
dm s , by applying a steady-state approximation to Ph3P
In ¼ 9:9 ꢃ 10 mol dm
s
) and kA ¼ 1:1 ꢃ 10 mol
3
ꢁ1
ꢂ
ꢂþ
electron transfer from AcrH to Ph3PO2 occurs to yield
ꢂꢁ
ꢂꢁ
þ
57
and O2 in Scheme 1. The concentration of O2 was esti-
AcrH and Ph3PO2. Ph3PO2 reacts with Ph3P to yield
Ph3P=O via O–O bond cleavage in competition with the dis-
sociation to Ph3P and O2 (Scheme 2) as the case of the photo-
ꢁ
9
ꢁ3
mated to be 9:1 ꢃ 10 mol dm by using Eq. 3,
O2ꢂꢁꢄ ¼ ½Ph3Pꢂþꢄ ¼ ðIn=kAÞ
assuming that the quantum yield for the radical coupling of
1=2
½
;
ð3Þ
þ
catalytic oxygenation of Ph3P with O2 catalyzed by Acr –Mes
in Scheme 1. The quantum yields of the AcrH -catalyzed pho-
þ

K. Ohkubo et al.
Bull. Chem. Soc. Jpn. Vol. 79, No. 10 (2006) 1495
þ
0
tooxygenation are always lower compared to that of Acr –
Mes (Fig. 1a), This is ascribed to the absence of the radical
(E ox) of benzylamine was determined by SHACV to be
0.62 V vs SCE (see Supporting Information Fig. S3a). Electron
ꢂþ
ꢂꢁ
ꢂ
ꢂþ
coupling reaction between Ph3P and O2 (Scheme 2). In
transfer from benzylamine to Acr –Mes is energetically fea-
0
þ
ꢂꢁ
the case of AcrH , no O2 is formed, because the electron-
47
sible, because the E red value of benzylamine is more negative
ꢂ
þ
ꢂ
ꢂþ
0
transfer reduction of O2 by AcrH (ꢁ0:43 V vs SCE) is en-
than that of Acr –Mes (E red ¼ 1:88 V vs SCE). On the oth-
ꢂꢁ
ꢂ
ergetically impossible. Thus, Acr –Mes is a better photocata-
þ
er hand, O2 is formed by electron transfer from the Acr
39
lyst than AcrH .
Photocatalytic Oxidation of Benzylamine. Visible light
moiety to O2. Consequently, proton transfer from benzyl-
ꢂꢁ
ꢂ
amine radical cation to O2 occurs to yield HO2 and the neu-
tral radical species.
þ
irradiation of the absorption band (ꢁ > 430 nm) of Acr –
ꢁ
3
ꢁ3
Mes (5:0 ꢃ 10 mol dm ) in an O2-saturated acetonitrile
Whether the neutral species derived from deprotonation of
benzylamine radical cation is a nitrogen-centered radical
2
(
MeCN) solution containing benzylamine (6:0 ꢃ 10ꢁ
ꢁ3
ꢂ
ꢂ
mol dm ) resulted in the formation of N-benzylideneben-
zylamine, PhCH2N=CHPh (Eq. 4).
(PhCH2NH ) or a benzyl type radical (PhCH NH2) has yet
to be determined (Scheme 3). The radical species has been di-
rectly detected by ESR in the hydrogen abstraction from ben-
zylamine by t-butoxyl radical in order to determine the struc-
2
PhCH2NH2 þ O2
hꢃ
65–67
ture of the hydrogen-abstracted radical (Scheme 4).
The photoirradiation of a degassed neat solution of
ꢁ
ꢁꢁꢁꢁꢁꢁꢁ! PhCH2N=CHPh þ H2O2 þ NH3: ð4Þ
þ
Acr {Mes
The yield of PhCH2N=CHPh and H2O2 was determined to be
1
5
Experimental).
8 and 46% by H NMR and iodometry, respectively (see
6
0.08
2,63
[PhCH NH ] = 0 mol dm–3
2
2
Laser flash photolysis was used to elucidate the mechanistic
details for photocatalytic oxidation. Transient absorption spec-
tra taken after the nanosecond laser excitation at 430 nm of a
0.06
0
0
.04
.02
þ
degassed MeCN solution of Acr –Mes in the absence and
presence of benzylamine are shown in Fig. 6. The absorption
ꢂþ
36
ꢂ
band due to the Mes moiety (ꢁmax ¼ 470 nm) of Acr –
ꢂþ
Mes was quenched by the addition of benzylamine. How-
0
ever, no transient absorption due to benzylamine radical cation
was observed by nanosecond laser flash photolysis, probably
because of the rapid decay of benzylamine radical cation (vide
[
PhCH2NH2]
–0.02
–0.04
= 5.0 x 10 mol dm–3
–4
6
4
infra). The bimolecular rate constant of electron transfer
ꢂþ
ꢂ
from benzylamine to Acr –Mes (ket) was determined from
ꢂþ
300
350
400
450
500
550
600
the decays of transient absorption due to the Mes moiety
Wavelength / nm
in the presence of various concentrations of benzylamine as
shown in Fig. 7. The decay at 470 nm obeys pseudo-first-order
kinetics and the pseudo-first-order rate constant increases line-
arly with increasing concentration of benzylamine (Fig. 7b).
The ket value was determined from the slope of Fig. 7b to be
Fig. 6. Transient absorption spectra observed in photoin-
duced electron-transfer oxidation of benzylamine (5:0 ꢃ
ꢁ
4
ꢁ3
þ
ꢁ5
ꢁ3
)
10 mol dm ) with Acr –Mes (8:0 ꢃ 10 mol dm
taken 10 ms after laser excitation at 430 nm in deaerated
9
ꢁ1
3
ꢁ1
3
:3 ꢃ 10 mol dm s . The one-electron oxidation potential
CH CN at 298 K.
3
(
a)
(
b)
0
.15
1
05[C6H5CH2NH2] / mol dm–3
0
0.04
0
0
1
2
.43
.87
.3
0
.03
.02
0.10
.6
0
0
.05
0
0.01
0
0
20
40
60
80
100
0
1
2
3
5
–3
Time /
µ
s
10 [benzylamine] / mol dm
þ
ꢁ5
Fig. 7. (a) Decay time profiles at 520 nm in the photooxidation of benzylamine in MeCN containing Acr –Mes (8:0 ꢃ 10
ꢁ3
mol dm ). (b) Plot of pseudo-first-order rate constant vs concentration of benzylamine.

1
496 Bull. Chem. Soc. Jpn. Vol. 79, No. 10 (2006) BCSJ AWARD ARTICLE
PhCH NH
PhCH NH
2
2
h
ν
−
−
e
t
t
t
·
(
1/2)Bu OOBu
Bu O
and/or
·+
−H+
PhCH NH2
PhCH NH2
2
2
PhCHNH2
PhCHNH2
t
PhCH NH2
Bu OH
2
Scheme 3.
Scheme 4.
(a) Exp.
g = 2.0037
(
c)
0
.349 mT
(
0.466 mT)
1
.371 mT
0
.104 mT
(
1.406 mT)
(
0.184 mT)
0
.104 mT
H
2
mT
H
C
(
0.125 mT)
H
NH2
(b) Sim.
H
0
.554 mT
0
.560 mT
(
0.563 mT)
(
0.469 mT)
t
t
ꢁ1
ꢁ3
Fig. 8. (a) ESR spectrum under irradiation of a deaerated Bu OOBu solution containing benzylamine (3:0 ꢃ 10 mol dm ) at
243 K. (b) The computer simulated spectrum. (c) Hyperfine coupling constants (the values in parentheses are those determined
by DFT calculation).
(a) Exp.
g = 2.0037
(
c)
0
.349 mT
(
0.466 mT)
1
.371 mT
0
.104 mT
(
1.406 mT)
(
0.184 mT)
2
mT
0
.016 mT
H
H
C
(
0.019 mT)
(b) Sim.
H
ND2
H
0
.554 mT
0
.560 mT
(
0.563 mT)
(
0.469 mT)
t
t
2
ꢁ1
Fig. 9. (a) ESR spectrum under irradiation of a deaerated Bu OOBu solution containing [N,N- H2]benzylamine (3:0 ꢃ 10
ꢁ3
mol dm ) at 243 K. (b) The computer simulated spectrum. (c) Hyperfine coupling constants (the values in parentheses are those
determined by DFT calculation).
t
t
ꢁ1
ꢁ3
Bu OOBu containing benzylamine (3:0 ꢃ 10 mol dm ) at
43 K with a 1000 W high-pressure mercury lamp resulted in
the hyperfine coupling constants (hfc) were determined,
Fig. 8c. The computer simulated spectrum using these hfc val-
ues (Fig. 8b) agrees with the observed ESR spectrum. Deute-
rium substitution of the two hydrogen atoms of the amino
2
formation of a hydrogen-abstracted radical, which was clearly
observed by ESR as shown in Fig. 8. The hydrogen-abstracted
radical is formed via hydrogen abstraction from benzylamine
2
group of benzylamine ([N,N- H2] benzylamine) confirmed
t
ꢂ
by t-butoxyl radical (Bu O ), which is generated by the homo-
the hfc assignment in Fig. 9 since the observed ESR spectrum
agrees with the computer simulated spectrum using the same
hfc values except for the value of the deuterium (I ¼ 1) at
the N-position. The hfc values of the hydrogen-abstracted radi-
t
t 65–67
lytic cleavage of the O–O bond of Bu OOBu .
2.0037) of the ESR signal in Fig. 8 is typical for that of a ben-
zyl radical. From the well-resolved ESR spectrum in Fig. 8a,
The g value
(

K. Ohkubo et al.
Bull. Chem. Soc. Jpn. Vol. 79, No. 10 (2006) 1497
PhCH NH
2
2
·
·+
Acr −Mes
hν
·+
PhCH NH
2
2
+
·
·
PhCHNH2 + HO2
Acr −Mes
Acr −Mes
PhCH=NH + H O
2
2
PhCH NH
2
2
·
O2
−
O2
PhCH=NCH Ph + NH3
2
Scheme 5.
g = 2.1266
g = 2.0033
g = 4.146
x 5
3
mT
2
0 mT
þ
ꢁ3
ꢁ3
)
Fig. 10. ESR spectra under photoirradiation of an O2-saturated benzylamine solution containing Acr –Mes (5:0 ꢃ 10 mol dm
at 123 K.
2
cal of [N,N- H2] benzylamine become smaller by the magne-
togyric ratio of proton to deuterium (0.153) as compared to
those of non-deuterated benzylamine.68
Table 1. Zero-Field Splitting Parameters (D, E), and Dis-
tance of Contact Ion Pair (r)
˚
r/A
D/mT
E/mT
The hfc assignment is also supported by the DFT calcula-
tion (see the calculated hfc values in parentheses in Fig. 9b).
Total energy of both hydrogen-abstracted radicals were
Triphenylphosphine
Benzylamine
7.0
4.4
0.69
0.37
7.4
8.6
ꢀ
also determined by DFT using the UB3LYP/6-31G basis
set (see Experimental). The energy of the benzyl radical is
firmed by ESR measurements. When an O2-saturated benzyl-
þ
ꢁ1
8
kcal mol lower than that of nitrogen-centered radical. This
amine solution of Acr –Mes was irradiated at 123 K, a triplet
ESR signal was observed as shown in Fig. 10. A small signal
indicates that that deprotonation of benzylamine radical cation
occurs at the ꢄ-position.
Based on the above results, the reaction mechanism of pho-
was also observed at g ¼ 4:0 due to the ꢁMS ¼ 2 transition.
The ‘‘ꢁMS ¼ 2’’ line in the region of g ¼ 4:0 is a diagnostic
þ
70–73
tocatalytic oxidation of benzylamine Acr –Mes with molecu-
þ
marker for the triplet state.
The zero-field splitting param-
ꢅ
eters D and E (¼0) were determined from the triplet signals
ꢂþ ꢂꢁ
lar oxygen is shown in Scheme 5. Photoirradiation of Acr –
Mes gives the electron-transfer state, which can act as an
oxidant for benzylamine and a reductant for molecular oxygen,
resulting in the formation of a benzylamine radical cation and
due to the radical ion pair (PhCH2NH2 and O2 ) and the
values are listed in Table 1. When PhCH2NH2 was replaced
by Ph3P, the ESR signal of the triplet contact radical ion pair
ꢂꢁ
ꢂþ
ꢂꢁ
superoxide anion (O2 ). Then, proton transfer from benzyl-
ꢂꢁ
between Ph3P and O2 was observed, and therefore, the
value of D and E were larger (Fig. 11). Since D depends on
the distance between two electrons with parallel spins, the
average distance of two spins can be evaluated from the values
of D, and the distances between the radical ions (r) are also
listed in Table 1.
amine radical cation to superoxide anion (O2 ) occurs to give
the ꢄ-hydrogen-abstracted benzylamine radical and hydroper-
oxyl radical (HO2 ). The benzylamine radical reacts with
ꢂ
ꢂ
HO2 to yield benzylideneamine (PhCH=NH) and hydrogen
peroxide (H2O2). No further oxygenation of PhCH2N=CHPh
0
ꢂþ
occurs because the E ox value of PhCH2N=CHPh (2.09 V vs
SCE: see Supporting Information Fig. S3b) is more positive
The r value of the radical ion pair of PhCH2NH2 and
ꢂꢁ ꢂþ ꢂꢁ
˚
˚
O2 (8.6 A) is smaller than that of Ph3P and O2 (7.4 A).
0
ꢂ
ꢂþ
ꢂþ
ꢂþ
than the E red value of Acr –Mes . Instead, PhCH=NH re-
acts with benzylamine to produce the final product, PhCH2N=
The optimized structures of Ph3P and PhCH2NH2 were
calculated by density-functional theory (DFT) using the
6
3,69
ꢀ
CHPh (Scheme 5).
UB3LYP functional and 6-31G basis set (See Experimental),
Detection of Radical Intermediates by ESR. The occur-
rence of photoinduced electron transfer in the photocatalytic
oxygenation reactions described above has also been con-
and are shown in Fig. 12. The spin distributions of radical cat-
ions are also shown in Fig. 12. The unpaired electrons of
Ph3P are localized on the P atom. In contrast, the unpaired
ꢂþ

1
498 Bull. Chem. Soc. Jpn. Vol. 79, No. 10 (2006) BCSJ AWARD ARTICLE
M. P. Sibi, Wiley-VCH, Weinheim, 2001, Vol. 2. b) H. Togo,
Advanced Free Radical Reactions for Organic Synthesis, Elsevier,
Amsterdam, 2004.
g = 2.0023
4
Free Radicals in Biology and Environment, ed. by F.
Minisci, Kluwer, Dordrecht, 1997.
M. Schmittel, M. K. Ghorai, Electron Transfer in Chem-
5
istry, ed. by V. Balzani, Wiley-VCH, Weinheim, 2001, Vol. 2,
pp. 5–54.
6 L. Eberson, Electron Transfer Reactions in Organic Chem-
istry: Reactivity and Structure, Springer, Heidelberg, 1987,
Vol. 25.
3
mT
7
S. Fukuzumi, Advances in Electron Transfer Chemistry,
ed. by P. S. Mariano, JAI Press, Greenwich, 1992, Vol. 2, p. 65.
S. Fukuzumi, Electron Transfer in Chemistry, ed. by V.
Balzani, Wiley-VCH, Weinheim, 2001, Vol. 4, pp. 3–67.
S. Das, V. Suresh, Electron Transfer in Chemistry, ed. by
V. Balzani, Wiley-VCH, Weinheim, 2001, Vol. 2, pp. 379–456.
Fig. 11. ESR spectrum under photoirradiation of an O2-
ꢁ
2
saturated CH3CN solution containing Ph3P (1:0 ꢃ 10
8
ꢁ
3
þ
ꢁ3
ꢁ3
mol dm ) and Acr –Mes (5:0ꢃ10 mol dm ) at 123 K.
9
(
a) (b)
1
482.
0
C. Amatore, A. R. Brown, J. Am. Chem. Soc. 1996, 118,
1
1
1
a) S. Fukuzumi, K. Ohkubo, T. Okamoto, J. Am. Chem.
Soc. 2002, 124, 14147. b) S. Fukuzumi, Y. Fujii, T. Suenobu,
J. Am. Chem. Soc. 2001, 123, 10191.
1
2
Photoinduced Electron Transfer, ed. by M. A. Fox, M.
Chanon, Elsevier, Amsterdam, 1988.
a) F. M u¨ ller, J. Mattay, Chem. Rev. 1993, 93, 99. b) M.
1
3
Mella, M. Fagnoni, M. Freccero, E. Fasani, A. Albini, Chem.
Soc. Rev. 1998, 27, 81.
1
4
a) M. Julliard, M. Chanon, Chem. Rev. 1983, 83, 425.
Fig. 12. Optimized structures and spin distributions of (a)
ꢂþ
b) F. D. Lewis, Acc. Chem. Res. 1986, 19, 401. c) U. C. Yoon,
P. S. Mariano, Acc. Chem. Res. 1992, 25, 233.
Ph3P and (b) benzylamine radical cation obtained from
ꢀ
DFT calculations at the UB3LYP/6-31G basis set.
1
1
5
6
G. J. Kavarnos, N. J. Turro, Chem. Rev. 1986, 86, 401.
E. R. Gaillard, D. G. Whitten, Acc. Chem. Res. 1996, 29,
ꢂþ
electrons of PhCH2NH2 are delocalized on nitrogen atom
2
92.
1
ꢂꢁ
and aromatic carbons. Thus, O2 interacts more strongly with
ꢂþ
7
S. Fukuzumi, H. Imahori, Photochemistry of Organic
ꢂþ
Ph3P than with to that of PhCH2NH2 . This may result in
Molecules in Isotropic and Anisotropic Media, ed. by V.
Ramamurthy, K. S. Schanze, Marcel Dekker, New York, 2003,
pp. 227–273.
18 S. M. Hubig, J. K. Kochi, Electron Transfer in Chemistry,
ed. by V. Balzani, Wiley-VCH, Weinheim, 2001, Vol. 2, pp. 618–
676.
19 a) J. K. Kochi, Angew. Chem., Int. Ed. Engl. 1988, 27,
1227. b) R. Rathore, J. K. Kochi, Adv. Phys. Org. Chem. 2000,
35, 193.
ꢂþ
ꢂꢁ
the shorter distance of the Ph3P –O2 pair as compared with
ꢂþ
ꢂꢁ
the PhCH2NH2 –O2 pair.
þ
In conclusion, Acr –Mes has been shown to act as an effi-
cient photocatalyst for the oxygenation of different types of
substrates via the reaction of substrate radical cations and
superoxide anion. Radical intermediates in the photocatalytic
oxidation of substrates with O2 were successfully detected
by laser flash photolysis and ESR measurements to clarify
the photocatalytic mechanisms.
20
979, 101, 5961.
S. Fukuzumi, K. Mochida, J. K. Kochi, J. Am. Chem. Soc.
1
2
2
1
2
G. J. Kavarnos, N. J. Turro, Chem. Rev. 1986, 86, 401.
D. G. Whitten, C. Chesta, X. Ci, M. A. Kellett, V. W. W.
This work was partially supported by a Grant-in-Aid
Nos. 16205020 and 17750039) from the Ministry of Educa-
(
tion, Culture, Sports, Science and Technology, Japan.
Yam, In Photochemical Processes in Organized Molecular
Systems, ed. by K. Honda, Elsevier, Amsterdam, 1991.
23 G. Pandey, In Topics in Current Chemistry: Photoinduced
Electron Transfer V, Springer-Verlag, Berlin, 1993, Vol. 168,
pp. 175–221.
Supporting Information
1
þꢀ
Fluorescence quenching of AcrH by Ph3P (S1) and electro-
chemical measurements (S2 and S3). This material is available
free of charge on the web at http://www.csj.jp/journals/bcsj/.
2
4
a) D. Mangion, D. R. Arnold, Acc. Chem. Res. 2002, 35,
2
1
97. b) T. Miyashi, H. Ikeda, Y. Takahashi, Acc. Chem. Res.
999, 32, 815.
References
2
5
a) U. C. Yoon, P. S. Mariano, R. S. Givens, B. W. Atwater,
1
J. Fossey, D. Lefort, J. Sorba, Free Radicals in Organic
Chemistry, Wiley, New York, 1995.
B. Giese, Radicals in Organic Synthesis: Formation of
In Advances in Electron Transfer Chemistry, ed. by P. S. Mariano,
P. S. JAI Press, Greenwich, CT, 1994, Vol. 4, pp. 117–205.
b) H. D. Roth, In Topics in Current Chemistry: Photoinduced
Electron Transfer IV, ed. by J. Mattay, Springer-Verlag, Berlin,
1992, Vol. 163, pp. 131–245. c) F. D. Saeva, In Advances in
Electron Transfer Chemistry, ed. by P. S. Mariano, JAI Press,
2
Carbon–Carbon Bonds, ed. by J. E. Baldwin, Pergamon Press,
Oxford, 1986.
3
a) Radicals in Organic Synthesis, ed. by P. Renaoud,

K. Ohkubo et al.
Bull. Chem. Soc. Jpn. Vol. 79, No. 10 (2006) 1499
Greenwich, CT, 1994, Vol. 4, pp. 1–25.
2
49 The SHACV method provides a superior approach to
directly evaluating the one-electron redox potentials in the pres-
ence of a follow-up chemical and reaction, relative to the better-
known dc and fundamental harmonic ac methods. See: a) M. R.
Wasielewski, R. Breslow, J. Am. Chem. Soc. 1976, 98, 4222.
b) E. M. Arnett, K. Amarnath, N. G. Harvey, J. Cheng, J. Am.
Chem. Soc. 1990, 112, 344.
50 C. K. Mann, K. K. Barnes, In Electrochemical Reactions in
Non-aqueous Systems, Mercel Dekker, New York, 1970.
51 a) A. D. Becke, J. Chem. Phys. 1993, 98, 5648. b) C. Lee,
W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785. c) W. J. Hehre,
L. Radom, P. V. R. Schleyer, J. A. Pople, In Ab Initio Molecular
Orbital Theory, Wiley, New York, 1986.
6
a) S. Fukuzumi, In Advances in Electron Transfer Chemis-
try, ed. by P. S. Mariano, JAI Press, Greenwich, CT, 1992, Vol. 2,
pp. 67–175. b) S. Fukuzumi, S. Itoh, In Advances in Photochem-
istry, ed. by D. C. Neckers, D. H. Volman, John Wiley & Sons,
New York, 1999, Vol. 25, p. 107. c) A. Sancar, In Advances in
Electron Transfer Chemistry, ed. by P. S. Mariano, JAI Press,
Greenwich, CT, 1992, Vol. 2, pp. 215–272.
J. Lind, X. Shen, G. Mer e´ nyi, B. O. Jonsson, J. Am. Chem.
Soc. 1989, 111, 7654.
a) M. S. McDowell, J. H. Espenson, A. Bakac, Inorg.
¨
2
7
2
8
Chem. 1984, 23, 2232. b) K. Zahir, J. H. Espenson, A. Bakac,
J. Am. Chem. Soc. 1988, 110, 5059.
2
9
L. Eberson, R. Gonzales-Luque, J. Lorentzon, M. Merchan,
B. O. Roos, J. Am. Chem. Soc. 1993, 115, 2898.
E. G. German, A. M. Kuznetsov, I. Efremenko, M.
Sheintuch, J. Phys. Chem. A 1999, 103, 10699.
S. Fukuzumi, S. Fujita, Y. Suenobu, H. Yamada, H.
Imahori, Y. Araki, O. Ito, J. Phys. Chem. A 2002, 106, 1241.
a) C. S. Foote, Acc. Chem. Res. 1968, 1, 104. b) C. S.
52 A trace amount of Ph2P(=O)OPh (e.g., 0.6%) was formed
in the reaction of Ph3P with singlet oxygen. In the photocatalytic
oxygenation of Ph3P with Acr –Mes, no Ph2P(=O)OPh was
detected.
53 S. Tsuji, M. Kondo, K. Ishiguro, Y. Sawaki, J. Org. Chem.
1993, 58, 5055.
53
þ
3
0
3
1
3
2
54 M. Nakamura, M. Miki, T. Majima, J. Chem. Soc., Perkin
Trans. 2 2000, 1447.
Foote, E. L. Clennan, Properties and Reactions of Singlet Oxygen,
In Active Oxygen in Chemistry, ed. by C. S. Foote, J. S. Valentine,
A. Greenberg, J. F. Liebman, Chapman and Hall, New York,
ꢂþ
55 The formation of Ph3P was monitored at 380 nm in order
ꢂ
to avoid the decay of the absorption due to the Acr moiety at the
shorter wavelengths.
1
995, pp. 105–140.
a) D. R. Kearns, Chem. Rev. 1971, 71, 395. b) L. M.
Stephenson, M. J. Grdina, M. Orfanopoulos, Acc. Chem. Res.
980, 13, 419.
a) Singlet Oxygen Reactions with Organic Compounds and
3
3
56 G. J. Kavarnos, Fundamentals of Photoinduced Electron
Transfer, Wiley-VCH, New York, 1993.
57 D. G. Ho, R. Gao, J. Celaje, H.-Y. Chung, M. Selke,
Science 2003, 302, 259.
1
3
4
Polymers, ed. by B. Ranby, J. F. Rabeck, Wiley, New York, 1978.
b) Singlet Oxygen, ed. by A. A. Frimer, CRC Press, Boca Raton,
FL, 1985, Vols. I–IV. c) A. Griesbeck, In CRC Handbook of
Organic Photochemistry and Photobiology, ed. by W. M.
Horspool, P.-S. Song, CRC Press, Boca Raton, FL, 1995, p. 301.
58 Z. B. Alfassil, P. Neta, B. Beaver, J. Phys. Chem. A 1997,
101, 2153.
59 S. Yasui, S. Tojo, T. Majima, J. Org. Chem. 2005, 70,
1276.
60 R. Battio, H. L. Clever, C. L. Young, IUPAC Solubility
Data Series, Pergamon Press, Oxford, 1981, Vol. 7.
3
A. G. Greisbeck, J. Mattay, Marcel Dekker, New York, 2005,
5
M. A. Iesce, Synthetic Organic Photochemistry, ed. by
þ
61 The AcrH -catalyzed photooxygenation of substrates have
pp. 299–363.
already been studied extensively.47 See: a) K. Suga, K. Ohkubo,
S. Fukuzumi, J. Phys. Chem. A 2003, 107, 4339. b) K. Suga,
K. Ohkubo, S. Fukuzumi, J. Phys. Chem. A 2005, 109, 10168.
c) K. Suga, K. Ohkubo, S. Fukuzumi, J. Phys. Chem. A 2006,
110, 3860. d) K. Ohkubo, K. Suga, S. Fukuzumi, Chem. Commun.
2006, 2018.
62 No formation of PhCH2N=CHPh was observed in AcrH -
catalyzed photooxidation since the photocatalyst was decomposed
by the formation of a large amount of the adducts.
3
6
S. Fukuzumi, H. Kotani, K. Ohkubo, S. Ogo, N. V.
Tkachenko, H. Lemmetyinen, J. Am. Chem. Soc. 2004, 126, 1600.
K. Ohkubo, H. Kotani, S. Fukuzumi, Chem. Commun.
005, 4520.
H. Kotani, K. Ohkubo, S. Fukuzumi, J. Am. Chem. Soc.
004, 126, 15999.
3
7
2
2
4
3
8
þ
39
265.
K. Ohkubo, T. Nanjo, S. Fukuzumi, Org. Lett. 2005, 7,
4
0
Tetrahydrofuran as a solvent of the Grignard reaction is
63 F. D. Lewis, T.-I. Ho, J. Am. Chem. Soc. 1977, 99, 7991.
64 a) C. G. Shaefer, K. S. Peter, J. Am. Chem. Soc. 1980, 102,
7567. b) F. D. Lewis, Acc. Chem. Res. 1986, 19, 401.
65 P. J. Krusic, J. K. Kochi, J. Am. Chem. Soc. 1968, 90, 7155.
66 a) J. K. Kochi, P. J. Krusic, D. R. Eaton, J. Am. Chem. Soc.
1969, 91, 1877. b) P. J. Krusic, J. K. Kochi, J. Am. Chem. Soc.
1969, 91, 3938. c) P. J. Krusic, J. K. Kochi, J. Am. Chem. Soc.
1969, 91, 3942. d) J. K. Kochi, P. J. Krusic, J. Am. Chem. Soc.
1969, 91, 3944. e) J. A. Howard, E. Furimsky, Can. J. Chem.
1974, 52, 555.
67 a) S. Fukuzumi, K. Shimoosako, T. Suenobu, Y.
Watanabe, J. Am. Chem. Soc. 2003, 125, 9074. b) H. Kitaguchi,
K. Ohkubo, S. Ogo, S. Fukuzumi, J. Am. Chem. Soc. 2005, 127,
6605. c) K. Ohkubo, Y. Moro-oka, S. Fukuzumi, Org. Biomol.
Chem. 2006, 4, 999.
3
8
better than dichloromethane reported previously.
S. Fukuzumi, S. Koumitsu, K. Hironaka, T. Tanaka, J. Am.
Chem. Soc. 1987, 109, 305.
K. J. Smith, E. D. Bergbreiter, M. Newcomb, J. Org.
Chem. 1985, 50, 4549.
4
1
4
2
4
4
3
4
R. D. Mair, A. J. Graupner, Anal. Chem. 1964, 36, 194.
S. Fukuzumi, S. Kuroda, T. Tanaka, J. Am. Chem. Soc.
1
985, 107, 3020.
S. Fukuzumi, M. Ishikawa, T. Tanaka, J. Chem. Soc.,
Perkin Trans. 2 1989, 1037.
C. G. Hatchard, C. A. Parker, Proc. R. Soc. London, Ser. A
956, 235, 518.
K. Ohkubo, K. Suga, K. Morikawa, S. Fukuzumi, J. Am.
Chem. Soc. 2003, 125, 12850.
A. J. Bard, L. R. Faulkner, Electrochemical Methods:
Fundamentals and Applications, John Wiley & Sons, New York,
001, Chap. 10, pp. 368–416.
4
5
4
6
1
4
7
4
8
68 ESR measurements of the hydrogen-abstracted radical
of benzylamine in Figs. 8 and 9 were carried out under same
experimental and instrumental conditions i.e., concentration,
2

1
500 Bull. Chem. Soc. Jpn. Vol. 79, No. 10 (2006) BCSJ AWARD ARTICLE
temperature, and modulation width.
a) S. Naya, H. Ohtoshi, M. Nitta, J. Org. Chem. 2006, 71,
Elementary Theory and Practical Applications, McGraw-Hill,
New York, 1972.
72 S. Fukuzumi, T. Okamoto, K. Ohkubo, J. Phys. Chem. A
2003, 107, 5412.
6
76. b) S. Naya, J. Nishimura, M. Nitta, J. Org. Chem. 2005, 70,
9
1
9
780.
7
0
F. D. Lewis, T.-I. Ho, J. T. Sinpson, J. Am. Chem. Soc.
982, 104, 1924.
J. E. Wertz, J. R. Bolton, In Electron Spin Resonance,
73 T. Da Ros, M. Prato, D. M. Guldi, M. Ruzzi, L. Pasimeni,
Chem. Eur. J. 2001, 7, 816.
1
7
1